Authors: Mary Kaltenberg*, UNU-MERIT and The MIT Media Lab, The Collective Learning Group, Dominik Hartmann, University of Leipzig and University of Hohenheim, Cristian Jara-Figueroa, The MIT Media Lab, The Collective Learning Group
Topics: Economic Geography
Keywords: labor markets, education, regional development, education
Session Type: Paper
Start / End Time: 12:40 PM / 2:20 PM
Room: Regent, Marriott, River Tower Elevators, 4th Floor
Presentation File: No File Uploaded
Recent methods in economic geography and complexity research predict the adoption of new industries in a region. However, may these methods also help predict regional occupational dynamics? Here we use a fine-grained dataset of formal workers in 558 micro-regions in Brazil to introduce (1.) a labor mobility space that connects occupations by the likelihood of the job switches between them and (2.) an industry-occupation space that connects occupations by the industries they share and thus creates a link between the industrial structure and the labor market. Our regressions show that the density of related occupations in the industry-occupation space and the density in the labor mobility space are significant predictors of regional occupational dynamics. They remain significant, even after controlling for occupational diversification trends within the regions, occupational diversification trends in neighboring regions, demand replacement factors, population size and educational differences. Finally, our methods allow also to revealing the comparative over- and underrepresentation of occupations in the Brazilian micro-regions according to their product structures. Our results imply that labor market predictions need to consider the unique structural constraints and opportunities of different regions, and can make use of new occupational relatedness measures, to predict regional occupational dynamics.